"Presidebilismo" In South America - Illinois State University Websites

advertisement
“PRESIDEBILISMO” IN SOUTH AMERICA:
THE ARGENTINEAN CASE
Christopher A. Martinez
Loyola University Chicago
Abstract:
In the last 27 years, seventeen presidents throughout the
world have not been able to complete their constitutional term in
office. But more surprisingly, twelve of them took place in South
America where presidents enjoy strong formal powers. This is the
paradox of presidebilismo: presidents facing high risk of early departure
from office in countries where they are endowed with strong
constitutional powers vis-à-vis the legislature1. By using a path
dependent approach, I look at four presidencies in Argentina from
1983 until 2001 to examine the phenomenon of presidebilismo. I
analyze the two ill-fated administrations of Raul Alfonsín (1983-1989)
and Fernando de la Rúa (1999-2001), and also include Carlos
Menem’s administrations (1989-1995 and 1995-1999), who
completed two terms in office. Findings suggest that executivelegislative relations and economic conditions act as major structural
determinants of presidebilismo, while mass mobilization works as a
“precipitant.” Alfonsín and De la Rúa’s early departures were
explained by the combination of minority government and serious
economic woes. In turn, Menem’s first term represents the opposite:
a not-threatening political and economic environment that enabled
him even to seek reelection. Menem’s second period shows how a
majority government canceled out the effects of a worsening
economy and social protests allowing him to stay in office.
1
I want to thank to Dr. Peter Sanchez (Pol. Science Dept. at Loyola University
Chicago) who came up with the term presidebilismo, which combines two
Spanish words “presidencialismo” (presidentialism) and “débil” (weak).
1
Keywords: presidentialism - presidential breakdown –
Argentina – executive-legislative relations
I)
Introduction
South American countries have enjoyed stable and
uninterrupted democracies in the last 20 years. While the democratic
transition stage appears to be completed, democratic consolidation
still requires more attention. Democratic stability, therefore, cannot
be automatically interpreted as government stability. Indeed, several
presidents have been removed from office or have prematurely
ended their term. In the last 27 years, twelve out of seventeen
presidential breakdowns throughout the world have taken place in
South America. While some South American presidents have been
overthrown by military means, others have voluntarily resigned or
been removed by institutional mechanisms. This is a unique
phenomenon especially considering that South American countries
have historically been characterized by giving strong constitutional
powers to the president vis-à-vis the legislature. In fact, one key
feature of Latin American political systems has been presidencialismo,
or the overbearing power of the chief executive vis-à-vis the
legislatures and courts. We are thus left with the puzzle of why it is
that many presidents have not been able to complete their term in
office in several South American countries? This phenomenon can be
depicted as presidebilismo: presidents facing high risks of presidential
breakdown even though they are endowed with strong formal
powers. Hence, the objective of this paper is to explain which factors
drive presidebilismo.
In the following section I will review the literature on
presidential failures by looking at the three major determinants:
institutional, economic, and social. In addition, I will describe how
these different approaches explain presidebilismo. Section III addresses
2
the operationalization of the explanatory model that I propose; the
conceptualization of the study’s variables, the selection of cases, and
the method I will employ. In section IV, I will analyze the
presidencies of the cases that I selected by looking at how executivelegislative relations, economic conditions, and mass mobilizations
interact to affect presidebilismo. In the final part of this work, I present
the conclusions, limitations and needs for further research.
II)
Literature review: the causes of Presidebilismo
Even though the premature termination of a presidency is
not a recent phenomenon, the scholarship has only recently started to
address this phenomenon. Most of these works have focused on the
causes or processes that led to such outcomes, the different types of
premature presidential endings, and their consequences. Whereas
some research has looked at the problem of presidential instability
from a broader perspective, including countries from more than one
geographical region (Baumgartner and Kada 2003; Kim and Bahry
2008; Pérez-Liñán, 2007); others have paid closer attention to Latin
America and South America due to their high number of interrupted
presidencies (Hochstetler 2006; Hochstetler and Edward, 2009;
Llanos and Marsteintredet 2010; Negretto 2006; Valenzuela 2004).
Institutional determinants
The issue of presidential breakdown has been related to the
so-called inherent instability of presidential regimes (Alvarez and
Marsteintredet 2010). In comparing presidential and parliamentary
regimes, Linz (1990, 65) argues that even when both systems face
executive crisis, in parliamentary systems this would become a
“government crisis,” whereas in presidential systems this would
constitute “regime crisis.” This is due to the dual democratic
legitimacy in presidential systems, where the president (who is,
simultaneously, head of government and the state) and the legislature
3
are popularly elected for fixed and independent terms (Linz 1994, 6).
Linz (1994) provides several examples where executive crisis in Latin
American countries later on turned out to become regime crisis
during the 1960s and 1970s. According to Linz (1994, 7), times when
executives and legislatures are deadlocked can be especially “complex
and threatening” in presidential systems for there is no democratic
principle or rule that can solve such a conflict.
Kim and Bahry’s (2008) work on interrupted presidencies in
third wave democracies shows that divided governments, where the
legislature is controlled by different parties or coalitions from the
executive; low first-round vote share; and party fragmentation
increase the odds of the president’s resignation or removal. Mustapic
(2010) points to the congress as the key institutional feature of
presidential democracy. She argues that the legislatures would allow
for institutional resolution of government crisis similar to what the
legislative branch does in parliamentary systems. Indeed, Llanos and
Marsteintredet (2010, 216-220) state that legislative action stands as
the “main force” driving a presidency’s premature end, which might
be derived from a president’s low share of votes in the legislature.
Negretto’s (2006) findings show that minority governments increase
the likelihood of presidential breakdown. Nonetheless, Cheibub’s
(2002) research on presidential democracies shows that
presidentialism is neither related to deadlocks, minority governments,
nor conflicts; thus, presidential breakdowns would not be affected by
these factors.
Pérez-Liñán (2007) states that presidents need a “legislative
shield” to protect them from social mobilization. This “shield”
requires support from a sufficient share of members of the
legislature, which in turns requires cohesive and disciplined parties.
Valenzuela (1994, 12) argues that Latin American presidents are
especially weak and constantly threatened by the existence of
“fragmented parties with little or no internal discipline.” He also adds
that, under minority governments, presidents usually experience
4
serious problems in generating “legislative support,” even from
his/her own party (Valenzuela 2004, 12). Indeed, in presidential
regimes there seems to be a lack of incentives for executive-party
cooperation. This is so since e.g. opposition’s politicians see strong
presidents as threatening to their own interests, and they also would
rather avoid relations with weak presidents (Valenzuela 2004, 13).
Mustapic (2010, 23) explains that in a two-party system it is less likely
that presidents would be backed up by the legislature, since the
opposition has no major incentives for sharing the “costs of the crisis
with the government party.”
Mainwaring and Shugart’s (1997) work brings up more
nuances to the discussion of institutional variables by assessing the
interaction between party system and presidentialism during times of
crisis. They find that the interaction between presidentialism and
party system also affects the level stability/consolidation of
democracy, where fragmented party systems and undisciplined parties
make the system more unstable. In addition, several researchers have
shown the difficulty for holding long-lasting party coalitions in
presidential regimes (Amorim Neto 2006; Deheza 1998; Thibaut
1998). However, this does not mean that long-lived coalitions are not
possible in presidential systems. For example, Chile and Brazil have
enjoyed successful coalitions since both returned to democracy
(Llanos and Marsteintredet 2010). These coalitions may be possible
since coalitions in presidential regimes work under two different
dynamics. Lijphart (1999, 105) explains that presidential regimes are
by definition minimal winning coalitions when it comes to the
presidency; but they also can be minimal winning, oversized, or
minority cabinets with regards to legislative support.
Economic factors
Others scholars have stressed the role of economic
performance on presidential breakdowns (Alvarez and Marsteintredet
2010; Hochstetler 2006; Kim and Bahry 2008; Llanos and
5
Marsteintredet, 2010; Negretto 2006). Presidential stability might be
threatened by high rates of inflation or slow economic growth. For
instance, Llanos and Marsteintredet (2010, 220) find that economic
growth proved to have a significant effect on presidential
breakdowns. They also show that a common feature of most
interrupted presidencies has been economic mismanagement by the
government, something with which Hochstetler and Edward (2009)
agree. Kim and Bahry (2008) show that a poor economic
performance increases the likelihood of presidents “leaving office
early.” Negretto (2006, 87) complements this argument by claiming
that a premature presidential term usually takes place amid social
mobilization triggered by “unpopular economic policies.”
In addition, Alvarez and Marsteintredet (2010, 49) posit that
both democratic and presidential breakdowns share some basic
determinants, e.g., economic performance and social mobilization. It
is worth noting Przeworski and Limongi’s (1997) findings about the
relationship between economic development and democratization.
They find that countries with a declining economy and with a per
capita income between $1,000 and $2,000 face more troubles when it
comes to democracy’s survival than countries with $1,000 or less but
with a growing economy (Przeworski and Limongi 1997, 177). This
might help to understand the challenges that South American states
may face when it comes to presidential instability. However, even
though democratic and presidential breakdowns may share some
common causes, their outcomes are fairly different. For instance, in
South America after 1978 all presidential breakdowns have been
followed by civilian governments, not military governments
(Hochstetler 2006). Likewise, Alvarez and Marsteintredet (2010)
point out that, even when presidential regimes are unstable, in the last
decades this instability has not translated into a regime failure, but
rather into a government crisis.
Social mobilization:
6
A final factor driving presidential instability is social
mobilization (Hochstetler 2006; Hochstetler and Edwards 2009; Kim
and Bahry 2008; Llanos and Marsteintredet 2010). People on the
street demanding the president’s ouster has usually been triggered by
political scandals and poor economic performance. Hochstetler
(2006) argues that social movements affect presidential stability
especially when linked with legislative actions. Hochstetler (2006,
409) defines this as a “mutually reinforcing collaborative action” that
has been able to successfully depose presidents from office in South
America. She also claims that when institutional means have not
found people’s support, motions for removing presidents have not
worked out; however, social mobilizations on their own have often
turned out successful. Kim and Bahry (2008) support Hochstetler’s
findings, and also point out the importance of social mobilization.
Kim and Bahry (2008) even state that in Latin America social
mobilizations have stronger effects on presidential stability when
compared with African countries. Additionally, Hochstetler and
Edwards (2009) show that high levels of perception of corruption
and the number of deaths as a result of government repression
against protesters have a significant effect on explaining the
failure/success of challenges to presidents.
III)
Research Design
The Model: two structural determinants and one precipitant
The major factors affecting presidential breakdowns
identified in the literature are institutional, which mainly stress the
executive-legislative relationships. Among non-institutional factors
are mass mobilization, poor economic performance, scandals in
which the president has been directly involved, and the number of
dead protesters. The relationship among these factors can be
understood either by following the explanations by Llanos and
Marsteintredet (2010); or by Hochstetler and Edwards (2009), and
7
Kim and Bahry (2008). On the one hand, the role of executivelegislative relations becomes more relevant for explaining the final
outcome. The existence of a unified, majority opposition may be
sufficient to depose a president from office when political and/or
economic crisis take place, regardless of the presence or strength of
mass mobilization. According to Hochstetler (2006), and Kim and
Bahry (2008), however, after political scandals or economic crisis
takes place, a president’s survival hinges on the degree of social
mobilization. Legislative action and street protests may support each
other for the purpose of deposing a president. Although, it may be
possible that mass mobilizations on their own can bring down a chief
executive. At Figure 1, we can see a potential causal diagram for
presidential breakdown. Under minority government situations the
opposition might be even less likely to support the government and
share the responsibility about political costs derived from economic
crisis and street protests’ repression.
Figure 1: Causal paths and interactions of Presidebilismo’s
determinants
Poor
Economic
Performance
Scandals
Minority
government
Presidebilismo
Mass
Mobilization
The diagram shows that under poor economic conditions or
high political scandals that involve the executive, presidents would be
at risk if his/her party or coalition does not have a majority in the
legislature and if this is accompanied by social mobilizations. In
addition, it is important to note that there is a closer relation between
economic crisis/scandals and mass mobilization than with minority
government. Mass mobilizations usually arise in response to poor
economic conditions. For instance, Corrrales (2002, 37) posits that in
8
less stable democracies when the economy performs poorly, people
mobilize in order to protest against the government and politicians
that (they believe) are corrupt, while exhibiting strong “antiestablishment sentiments.” Mass mobilizations then are the last visual
reaction of economic crises. Leaving out political scandals, street
protests seem unlikely to appear under good economic conditions.
Thus, an economic crisis becomes a necessary condition for the
upsurge of mass mobilizations, which can in turn affect the severity
and length of the crisis.
Minority governments, on the other hand are not necessarily
a consequence of economic crisis. It is true that a ruling party or a
coalition may be electorally punished for its economic
mismanagement, but that would not be the only determinant of a
minority government. Intra-party struggles, coalition breakdowns,
ideological confrontations, a government’s inability to deliver
electoral promises, executives that do not compromise with other
parties, and electoral rules are some of the possible reasons for the
existence of a minority governments. Whereas economic crisis and
mass mobilizations are closely related; economic crisis and minority
governments work under different dynamics.
Therefore, the model that I propose will look at the
likelihood of presidebilismo by focusing on economic conditions and
the existence of a minority government. Yet, this model implies that
more and stronger street protests will result as the severity of an
economic crisis increases. Mass mobilizations are understood as
“precipitants” or “triggers” of presidential early departure, which may
speed up presidential removal. As Figure 2 shows, the prospect for a
president’s survival is lowest under poor economic conditions and
minority government; whereas it is highest when the economy does
well and the executive’s party has a majority of seats in the legislature.
The interaction between economic crisis and minority government
sets the scenario within which non-governmental political actors will
9
become active. For example, civil society, student organizations, and
unions may become relevant actors under these structural conditions.
Figure 2: Interaction Economic Situation, Minority Government and
Presidential Ouster Risk
High
Presidential
Ouster Risk
Bad
Good
Economic Situation
Majority Government
Minority Government
Figure 3 shows how the two structural determinants interact
Low may increase or reduce presidential
and how these interactions
survival. For instance, when both structural conditions are met, in
quadrant C, prospects for presidential survival are at the minimum;
whereas in quadrant B, where neither structural condition exist, the
elected president has the highest chances to remain in office.
However, when only one structural condition exists, in the quadrants
A or D, they will tend to offset each other, raising the probability that
the president will complete his/her constitutional term as expected.
Consequently, I propose the following working hypotheses:
H1: Presidential ouster risk will be the highest under economic crisis and
minority
government.
H2: Presidential ouster risk will be the lowest under good economic
conditions and when
the president’s party or coalition
controls the legislature.
H3: During economic crisis the president may survive in power as far as
s/he enjoys a
majority government.
H4: The president may finish his/her constitutional term in office under
a minority
government if there is no economic crisis.
Figure 3: Four Scenarios for President Survival as results of
Economic Conditions/Executive-Legislative Relations Interactions
10
Majority
government
A
B
I
C
D
G
Exe-Leg
Bad L
H Good
Economic Situation
Operationalizing Presidebilismo:LThe case of Argentina
B
Minority
government
L
Drawing on previous works (Cheibub
2002; Hochstetler
i
2006; Kim and Bahry 2008; Negretto 2006), I define the concept of
presidebilismo as a political crisis where a president, who ghas been
popularly elected in competitive elections, is forced to leave
h office
early by means of impeachment, legislative initiatives, or resignation.
Presidebilismo describes the situation where presidents are unable to
fulfill his/her constitutionally term in office but without
compromising the democratic order. What distinguishes the concept
of presidebilismo from other concepts such as “interrupted
presidencies” or “presidential breakdowns” is that in the former the
presidential ouster is not an outcome of the use of force or its threat
thereof by either the armed or police forces of the country.
From a methodological stand, most works on this topic
suffer from the problem of limited variance in the dependent
variable; that is, only studying cases that have experienced
presidential breakdowns. The only exceptions are Hochstetler (2006),
and Hochstetler and Edwards (2009) who have allowed for some
degree of variation on the dependent variables by focusing on
“challenges” to presidents, which means that even if the attempt to
remove the president fails, it is included in the analysis. Most works
addressing this phenomenon have used cross-national and time-series
quantitative analysis; thus, there is a need for bringing more
methodological diversity that can improve our understanding of
causal mechanisms.
11
The quantitative-leaning tendency for addressing this topic
has overlooked how the independent variables interact to impact on
the dependent variable. Simple correlations found in quantitative
works do not provide sufficient information whether the timing
when independent variables take place matters or not. I will adopt a
path dependence method in order to comprehend how certain events
may affect events that take place at a later time (Mahoney and
Villegas 2006, 78). This is important since, for example, the effects of
policies or plans launched by one administration may still have far
reaching effects into the future, enhancing or threatening the stability
of incoming administrations. The same is true about economic
conditions: the range of maneuver of a newly elected president would
be conditioned by how the economy was and is doing when s/he
takes office. And, an election that results in a minority government
will undoubtedly affect a president’s ability to stay in office, especially
if the economy sours and people take to the streets.
In addition, due to the “useful variation” of presidebilismo is
rare, a case-study approach offers more opportunities for uncovering
causal mechanisms than medium or large N studies (Gerring 2006,
111). Even with the lack of generalizability, case studies can provide
significant insights about precise causal mechanisms. In fact, the real
number of cases of presidebilismo is not very high, so a mid-N study
can address a significant number of the total number of cases.
Therefore, using a path dependent method and a case-study
approach, this paper attempts to empirically analyze how the main
variables discussed in the literature review -executive.-legislative
relations, social mobilization, and economic crisis- affect the
likelihood of presidebilismo’s occurrence.
Why Argentina?
Most presidential breakdowns in the world have taken place
in South America, specifically, in seven countries. In fact, three
12
countries have experienced the majority of interrupted presidencies.
In Argentina, Bolivia and Ecuador presidents have failed to finish
their constitutional terms in office more than once in the last 30
years. Unlike Ecuador and Bolivia, Argentinean presidents have not
been deposed by indigenous mobilizations or military interventions.
Argentina represents a special case due to its homogeneous
population and high level of democracy and economic development
in comparison with Bolivia and Ecuador. Rather, economic crisis,
union-led street protests, and institutional arrangement have been
more important in explaining Argentinean presidebilismo. From the
1990s on, Argentina could rule out military influence on politics and
also was able to become in one of the most prosperous developing
countries in the world. Notwithstanding the ouster of two presidents,
the Argentinean democratic regime has not been threatened. The
Argentinean case also enables us to look closer at how economic
conditions and social mobilization combine with institutional
arrangements to affect presidential survival. As described in the
previous section, the scholarship is divided among those who regard
the executive-legislative relations as the main determinant of
presidential stability; those who stress social mobilization; and those
who focus on both factors.
Argentinean “presidebilismo” will be addressed by looking at
the country’s two deposed presidents, Raul Alfonsín (1983-1989) and
Fernando de la Rúa (1999-2001), who were popularly elected in
competitive elections. There were other Argentinean presidents who
were ousted from power as well (Adolfo Rodríguez Saá and Eduardo
Duhalde); but since they were appointed by Congress after De la
Rúa’s failure, they are not included in this study. In addition to
Alfonsín and De la Rúa's administrations, I will analyze Menem’s two
terms in office (1989-1995 and 1995-1999). I do this to add more
variation to the dependent variable “presidential stability.” Menem’s
administration will serve to see whether the same factors that affected
Alfonsín and De la Rúa ouster were present during his successful
terms in office.
13
Information about social protests is collected from Schuster’s
et al (2006) work that covers mass mobilizations in Argentina from
1989 until 2003. Although, data for Alfonsín’s entire administration
are not available, there is sufficient data for the period just prior to
Alfonsín resignation. In addition, I will use printed press information
to complement Schuster’s et al. (2006) data. Economic crisis
information is gathered from the World Bank data series. In order to
identify periods of economic downturns I look at the following
variables: GDP per capita, inflation rates, GDP growth, external debt,
and unemployment. All these data serve to compare the different
economic situations that the three presidents experienced.
Information about election results and legislative o congressional
party composition has been obtained from the Argentinean Ministerio
del Interior (Ministry of Internal Affairs. To complement this data, I
also refer to secondary data sources such as research articles
published in scientific journals that address the executive-legislative
relations in Argentina.
IV)
Data Analysis
Raul Alfonsín’s Presidency (1983-1989):
After several years under a military regime, Raul Alfonsín
took office in Argentina with the major goal of consolidating the
transition to democracy. He received an absolute majority of the
votes and became the first president from the Radical Party to win a
free and fair election since the 1940s. This electoral victory also
translated into the Chamber of Deputies where the Union Cívica
Radical (UCR, Radical Civic Union) obtained a slight majority of seats
(50.8%). However, Alfonsín’s party was not able to control the
Senate, where the UCR only got 39.1% of the seats, while the major
opposition party, the Partido Justicialista (PJ), held 45.7% of the seats.
Moreover, the PJ also won the majority of governorships, 54.5%
14
versus 31.8% of the UCR. Hence, from the very beginning, Alfonsín
suffered from a minority government.
It is worth noting that Argentinean parties are characterized
by weak party cohesion. Mustapic (2002, 27-28) states that there are
two major factors causing this: low ideological density and
decentralized structure (national, provincial, and local). The party
leader then relies mostly on his/her electoral potential (Mustapic
2000, 580). Mustapic (2000) argues that winning elections becomes
the decisive factor for obtaining party unity and cohesion, which
hinges around its successful leaders. Cheresky (1990, 54) even claims
that when Alfonsín ran for the presidency, he represented a minority
faction within the UCR; but he at least could count on strong public
support. Thus, Alfonsín led a minority government and a not-always
reliable party support, with no more tools than his “electoral success”
and public support.
The minority government situation meant that the Radicals
were not able to pass important legislation. Greater interparty
cooperation was hard to achieve given the internal struggles over the
PJ’s leadership as a result of its electoral defeat in 1983 (Cheresky
1990, 57). The historically close relationship between the PJ and the
sindicatos (unions) became evident during this period when the Central
General de Trabajadores (CGT, the General Workers Central) and its
leader, the moderate Peronist Saul Ubaldini, represented the “real”
opposition to the Alfonsín administration (ibid.). Indeed, the role of
the CGT in opposing Alfonsín was also significant for putting people
on the streets by organizing more than 13 general strikes (ibid.)
As soon as he took office, Alfonsín also had to deal with
harsh economic conditions inherited from the military government:
an inefficient and oversized bureaucracy (Llanos, 2010, p. 58); two
consecutive years with negative economic growth (1981=-5.7%, and
1982= -5.0%); an external debt that had increased more than five
times in the last seven years (Andreassi 2011), and hyperinflation
15
(344% in 1983). In 1985, with the situation not improving, one of the
first major strikes occurred. The CGT and Ubaldini were able to
gather around 200,000 people to protest the economic situation and
to ask for Alfonsín’s resignation (Montalbano 1985). The
government implemented the Austral Plan in an effort to minimize
the effects of the economic crisis and to calm social mobilizations.
The plan was very peculiar for it aimed to control inflation but
simultaneously to maintain the current low unemployment rates and
economic growth (Llanos, 2010, p. 58). The underlying idea of the
plan was to avoid negative consequences on people in order to not
put the democratic transition at risk (ibid.). The short term effects of
Plan Austral were positive though. The GDP grew dramatically from 7.59% in 1985 to 7.88% in 1986; while inflation dropped from 672%
to 90% in the same period (World Bank, 2011). The immediate
effects of Plan Austral were to avert a political crisis for Alfonsín. The
first mid-term election turned out positive for Alfonsín’s
government. Whereas the UCR kept its share of seats in the lower
house, the Partido Justicialista lost almost 5% of its seats.
Notwithstanding the PJ’s electoral defeat, nothing changed in terms
of Congressional majorities: the UCR kept control of the lower
house, while the PJ dominated/controlled the Senate.
At the end of 1986, Plan Austral’s positive effects proved to
be ephemeral. The economic situation became even worse. The
economy decreased and inflation increased, plus the business sector
and unions started to exert pressure over Alfonsín’s government
(Llanos 2010, 59). In response, the government launched the Plan
Primavera (Spring Plan) in August of 1987. Yet, it was too late for
increasing the UCR’s popularity. The second mid-term elections held
in September of 1987 left the government in a weaker position. The
consequences of the 1987-elections were important. First, the PJ got
74.3% of the governorships (a 22.8% increase), whereas the UCR was
able to retain only three out of 22 governors (in Cordoba, Rio Negro,
and the Federal District). Second, even though the PJ slightly
increased its share of seats in the Chamber of Deputies (from 40.6%
16
to 41.7%), the UCR’s share of seats declined to 44.7%, losing the
majority in the lower house and becoming the first minority party.
Third, the chances of a constitutional reform that would allow
Alfonsín to be reelected finally vanished (Llanos, 2010, p. 59).
At this point, the extreme scenarios described in the previous
sections – minority government and economic crisis with social
mobilization - were about to meet and combine. In this new and
more adversarial political scenario, Alfonsín’s options were drastically
reduced. The minority government situation, in which his party at
least controlled the Chamber of Deputies, had come to an end. Now,
the executive faced a “no majority” situation in Congress (Negretto
2006, 84). The economic woes resembled what happened in 1985,
but in 1989 Alfonsín had fewer political resources to rely on. For
instance, Alfonsín was not able to increase public spending since bills
on tax increases and privatizations were not approved in Congress
(Llanos 2010, 60). Furthermore, the economic crisis in 1989 was even
worse than the one in 1985. The dollar went up and reserves were
almost exhausted (Llanos 2010, 60); the GDP per capita declined from
$3,977 in 1988 to $2,380 in 1989; two years of negative economic
growth (-2.6% in 1988, and -7.5% in 1989); and inflation skyrocketed,
reaching almost 40% monthly in the first semester of 1989.
The only missing element was civic action, that is, the social
and political actors that will take advantage of the two interacting
structural conditions. The number of street protests went up to
almost 300 in the first six months of 1989, one of the highest levels
in Argentina’s recent history (Schuster et al. 2006, 30). Unions
organized 71% of the street mobilizations, and only a 13% of them
were led by neighbors and students (Schuster et al. 2006, 36). This
high level social mobilization created a political opening for the
opposition. The Partido Justicialista could benefit greatly from the
sindicatos’ actions by weakening Alfonsín’s government and also the
Radical candidate who was running for the presidency in 1989. In
fact, Llanos (2010, 61) points out that Peronist right-wings factions
17
might have instigated the looting that later on took place in May
1989.
The economic crisis, social mobilizations, and the
institutional paralysis resulted in a serious political crisis: the
resignation of Alfonsín’s economic minister and his team in March
1989 (Llanos 2010, 60). The appointment of a new economic
minister did not bring more consensus as expected, and eventually
the Peronist candidate, Carlos Menem, won the presidential election in
May 1989. The increasing street protests, looting, and violence that
left 15 people dead, set in motion Alfonsín’s early departure from
office. Alfonsín started negotiations with the elected president, Carlos
Menem, so as to agree on the terms of the anticipated handover.
Menem finally accepted Alfonsín’s offer and also secured the
approval of economic legislation in the lower house, where the
Radicals were the first majority party, until the new Congressional
elections in December 1989 (Llanos 2010, 61). Alfonsín resigned on
June 12th (five months early) during one of the most severe
economic, social and political crises that Argentina had experienced
in 30 years.
Even though Alfonsín led a minority government (from
1987 to 1989 he was under a divided government) and faced poor
economic conditions, his survival in office was only at risk when
major social protests arose. Yet, it is important to note that the
harshness of the economic woes was partially overcome by the
Austral and Primavera plans. Still, their effects were short-lived.
Alfonsín’s party losing its legislative majority and the economic crisis
worsening finally coalesced in 1989. Mass mobilizations took place,
demanding better salaries and a lower inflation, which put the
government in a weak position. When the president leads a minority
government, it is not only that the congress cannot provide a
“legislative shield” to deal with social mobilizations; but the
legislature might even leave the president with no political support,
making it difficult to survive a serious political and economic crisis.
18
In this case, not having a majority government, along with the severe
economic crisis and resultant social mobilization, led Alfonsín to
voluntarily leave the presidency before his constitutional term was
over.
Carlos Menem’s Presidency (1989-1999)
Even though Menem took office amid the worst
hyperinflationary crisis since Argentina returned to democracy in
1983, the president could rely on the electoral support he received in
the 1989 election, and also on the idea that the early start of his
presidency was necessary to save the economy. Furthermore, Menem
was also counting on the Justicialista majority in the Senate (56.5%)
and governorships (77.3%). However, the PJ did not control the
Chamber of Deputies where it became the first minority party
(44.1%) after the Congressional elections held on December 1989. In
turn, the UCR share of seats kept falling. In the lower house, the
UCR controlled only 35.4% of seats and 30.4% in the Senate.
Therefore, even though the economic situation might threaten
Menem’s government, he could still count on the “legislative shield,”
and most importantly he had ensured that major economic reform
would be backed by the opposition. Menem definitely had more
political and institutional resources to use at the start of his
administration than Alfonsín did during his whole term.
Yet, executive-legislative relations were not as smooth as
expected owing to congress’s composition. Ferreira and Goretti
(1998, 35) state that Menem faced opposition in Congress especially
after he started to issue decrees of necessity and urgency (DNUs).
His argument rested upon the need for speeding up law-making as
the situation required (ibid.). Nevertheless, the DNUs approval
required judicial revision, and since the Supreme Court’s members
were appointed under Alfonsín’s administration, Menem was not able
to ask for their removal. Instead, what he did was to pass a “packing
plan” in the Senate, where the PJ had an overwhelming majority, so
19
as to increase the number of members of the Supreme Court
(Ferreira and Goretti 1998, 36). In doing so, Menem paved the road
for issuing DNUs even during times of neither necessity nor urgency.
In his first term in office, Menem used 336 DNUs compared with 25
that had been issued between 1853 and 1989 (Ferreira and Goretti
1998, 33).
The goal of consolidating the democratic transition that
characterized Alfonsín’s government was of secondary importance
for Menem and citizens, principally because of the declining
economy. Even though Menem was elected on the basis of his
populist promises of salariazo (increasing salaries), as soon as he took
office, he carried out the most extensive neoliberal reform package
ever implemented in Argentina. In August 1989, Menem, supported
by the Radicals, enacted the State Reform Law, which basically
allowed for the privatization of most state assets (Ministerio del
Interior de Argentina 2008, 157-158). People came to believe that
state intervention in the economy was responsible for the egregious
economic outcomes of the Austral and Primavera plans (Ferreira and
Goretti 1998, 38). Consequently, when Menem took office, people
were much more concerned with reducing inflation and the state’s
role in the economy (ibid.). These demands were incorporated by
Menem when adopting several market-oriented policies, such as
deregulation, privatization, and the reduction of government
spending. These steps were taken because of the increasing
concentration of power in the executive branch (owing to the
DNUs), the Delegation Act authorized by Congress (controlled by
the Justicialistas), the “packing plan,” and the constant request by
citizens for fast solutions (ibid.).
In 1991, the economic situation significantly improved. The
GDP per capita went up to $5,733 (2,381 in 1989) being higher than at
any point during Alfonsín’s administration; economic growth reached
a historic double-digit number (12.67%)); and the major issue in the
last decade, inflation, finally seemed to be under control (172% in
20
1991 compared with 3,080% in 1989). While large-scale privatizations
were mostly responsible for increasing state revenues; the
Convertibility Law, which pegged the peso to the dollar (one-to-one),
was responsible for controlling the skyrocketing inflation. In 1994,
the annual inflation rate was 4.2%, economic growth hinged around
6%, and the GDP per capita increased to $6,967. On the other hand,
the unemployment rate, which historically had been low in Argentina
(one digit), started to go up in 1994 (12.1%). Nevertheless, the period
1991-1994 did not witness a significant number of mass
mobilizations (Schuster et al. 2006, 31). This may have been because
people preferred low inflation rates over full employment, therefore,
exhibiting “higher tolerance for recession” (Schamis 2002, 84).
In electoral terms, the PJ strengthened its dominance in the
Senate after the1992 mid-term elections by controlling 58.3% of
seats. In 1993, the Justicialistas also achieved an almost absolute
majority in the Chamber of Deputies (48.2%). With a near-majority
in the Congress and the governorships, and an economy that finally
took off, Menem’s stability in office was not under discussion. The
two major structural determinants of presidebilismo provided a secured
political, economic and social environment not only for Menem’s
survival in office, but also for his reelection. In 1995, Menem was
reelected thanks to the “Pact of Olivos,” between the PJ and formerpresident Alfonsín, as leader of the Radicals, which reformed the
constitution in 1994. Menem’s second term began with the
Justicialista’s electoral victory in the mid-term election held in 1995.
The PJ obtained an absolute majority in both the lower house
(51.4%) and the Senate (55.5%). The undesirable consequences of a
minority government were gone.
However, the economy started to exhibit some setbacks. The
GDP per capita had stagnated just above $7,000; economic growth
dropped to -2.5% in 1995; and unemployment reached 18.8%. State
revenues from privatization proved not to be long-lasting. For
instance, the privatization of the social security system began to
21
require more resources than the government possessed. The social
security deficits led the government to a “spiral of borrowing” since
as the public systems of social security had lost “contributors,” the
government still had to deal with “beneficiaries” (Levitsky and
Murillo 2003; Schamis 2002).
Public finances became more limited at the same time that
the executive had fewer policy tools to react under new economic
woes. The Convertibility Law that had successfully wiped out
inflation now appeared to be limited as a vehicle for economic
stimulation. The Convertibility Law required that the Central Bank
had to back all the Argentinean monetary base with international
reserves, which meant that the government could inject money into
the economy only when international reserves increased via “trade
surpluses or net capital inflows” (Schamis 2002, 83). No mechanisms
for economic stimulus remained under government control (Schamis
2002, 82). Basically, the Convertibility Law eliminated the possibility
of the executive using monetary and exchange policies to counteract
“economic shocks” (Levitsky and Murillo 2003, 153).
Social mobilizations were not absent during Menem’s
administration. Indeed, his government faced one of the largest
number of street protests in the period 1989-2003. The rising
unemployment and the declining economy were the driving forces
for the protesters. In spite of the high number of street protests,
especially in 1997, Menem’s survival in office was not affected.
During that period, Menem’s party enjoyed a majority within
Congress that served as a counterweight to the effects of street
protests. It is worth noting, though, that the main difference between
the mass mobilizations that took place under Alfonsín’s
administration and Menem’s second administration was the source of
their organization. Whereas 71% of the street protests under
Alfonsín were organized by sindicatos; during Menem’s second term
only 36% of them were union-led (Schuster et al. 2006, 38).
22
Coincidence or not, about 25% of Justicialista deputies were also labor
leaders during the 1990s (Corrales 2002, 34).
Internal struggles took place within the PJ when Menem
announced that he was seeking a third presidential term. Despite the
fact that a third term was unconstitutional and that 80% of the
population opposed to it, Menem thought that as long as his party
backed him, the Supreme Court would not be an issue (Corrales
2002, 33). Still, Justicialista leaders that considered Menem’s third term
bid as a threat to their own political careers rejected Menem’s second
reelection (Corrales 2002, 33). The PJ’s intra-party conflict
manifested itself in a spending race, which led to a rising national
debt (Corrales 2002, 32). As a matter of fact, in 1998 Argentina’s
external debt was $75,524 million, a 45.7% increase from 1989.
Menem’s populist message reemerged and, instead of cutting
spending, he proposed to increase the 1998 budget (ibid.). Finally, the
internal resistance against Menem’s desires for a third term was
successful, deterring him from seeking a second consecutive
reelection. Eduardo Duhalde, one of the strongest opponents to
Menem’s third period, was chosen as the presidential candidate for
the 1999 election. Whereas democratic principles prevailed; the fiscal
irresponsibility of Menem and provincial authorities put the economy
at risk, “increasing… [its] vulnerability to the external shock of 1998”
(Corrales 2002, 34).
Menem’s second term enables us to understand how the two
structural determinants of presidebilismo –a minority government and
economic woes- may affect a president’s prospects for survival.
During Menem’s first term the presidential stability was not at stake
even when inflation was not under control in 1989-1991. Legislative
strength, the growing economy, and the lack of significant mass
mobilization led Menem to his successful presidency and his
subsequent reelection. The second term shows that even when one
of the structural determinants turns out threatening to presidential
stability the other can act as a counterweight. Although the economy
23
declined, Menem still had a majority in the legislature during his
second term.
Fernando de la Rúa’s Presidency (1999-2001)
Fernando de la Rúa became the second Radical president,
after Alfonsín. The successive electoral defeats suffered during the
1990s had weakened UCR’s popular support. Therefore, so as to
defeat the PJ in the 1999 election, the UCR formed the “Alianza”
(Alliance) with the newly created conglomeration “Frente Pais
Solidario” (FREPASO, Country in Solidarity Front). The ideological
diversity of FREPASO, composed by some leftist and former
Peronistas groups, was mostly recognized by its two major leaders
Carlos “Chacho” Alvarez and Graciela Fernandez Meijide (Llanos
2010, 62). The potential of the Alianza rested upon the combination
of each partner’s strengths. FREPASO’s lack of well-developed
internal rules and structure could be overcome by the UCR’s
“national organizational structure” (Llanos and Margheritis 2006, 81).
On the other hand, FREPASO’s leaders enjoyed broad popular
support and media communication skills (ibid.), which were one the
major flaws of the UCR. The Alianza promised an end to corruption,
increased transparency, and increased investments in social sectors
(Corrales 2002, 34), while keeping the Convertibility Plan, and
improving “democratic governance mechanisms” (Llanos 2010, 62).
The Alianza was successful since Fernando de la Rúa easily
defeated the Justicialista candidate in the 1999 presidential election.
The Alianza was the first coalition to take office in recent
Argentinean history (Llanos and Margheritis 2006, 78). Still, the
Alianza’s popular support did not translate into the Congress. Despite
the Alianza’s advantage over the PJ (about 1% in the share of seats)
in the lower house, it was able to control only 39.7% of the seats,
becoming the first minority party. The PJ, in turn, had the majority in
the Senate (55.5% of seats) and governorships (62.5%). Again, the
24
same unfavorable legislative situation faced by Alfonsín’s
administration emerged in the De la Rúa government.
Soon after taking office, the coalition started to evidence
signs of divisions. De la Rúa’s relations with FREPASO became
complicated since the president only gave secondary government
positions to its leaders (Schamis 2002, 87). More surprisingly, De la
Rúa excluded influential Radical leaders from his inner circle, which
was formed by friends and family (e.g. his son and his brother) and
only a few young Radicals who had little influence within the party
(Llanos and Margheritis 2006; Schamis 2002). De la Rúa not only
faced an unfavorable environment in Congress, but now he had also
limited the political resources he might borrow from the two partners
of the Alianza.
Furthermore, the executive also had problems when
negotiating with the opposition. Llanos and Margheritis (2006, 89)
point out that the PJ was unwilling to cooperate with a “weak
executive,” and then share the costs of adopting unpopular policies.
The Justicialistas opposed De la Rúa’s bills in the Senate, especially
those involving labor issues. In fact, the PJ was not willing to
approve any labor reform under Menem’s administration.
Nevertheless, De la Rúa achieved a twofold victory in this area. His
bill on labor reform was finally approved, although resisted by his
coalition fellows in the lower house, and then passed in the Senate,
thus breaking the Peronista majority (Llanos and Margheritis 2006, 8990). This law became the major political victory of De la Rúa in a
clearly unfavorable legislative context. Yet, the initial excitement was
quickly replaced by concern. The Alianza became involved in a
political scandal as the government was accused of bribing senators
so as to get the labor reform law passed (Llanos and Margheritis
2006, 89-90). The Vice-President and FREPASO leader, Carlos
“Chacho” Alvarez, requested an immediate investigation of the
accusation. Since the government failed to investigate, in October
25
2000, Alvarez resigned, generating the first political crisis within the
executive (Levitsky and Murillo 2003, 154).
When De la Rúa took office the economy was already in
trouble. In addition to De la Rúa’s political isolation, the country
experienced high unemployment rates, negative economic growth,
and an increasing external debt. Alvarez’s resignation did nothing but
increase Argentina’s financial risk rate owing to fear of political and
financial instability (Schamis 2002, 87). After this, the risk of financial
default rose which made the situation for De la Rúa even worse
(Llanos and Margheritis 2006, 93). The Convertibility Plan had
already taken away the most important tools for dealing with
economic crisis from the executive. De la Rúa attempted to address
the economic woes via issuing DNUs and asking the Congress for
discretionary power for his economy minister (Schamis 2002, 88).
This not only did little to improve the economy, but also exacerbated
De la Rúa’s political isolation (Llanos and Margheritis 2006, 91).
External economic aid became urgent. However, external
funds, specifically from the IMF, were conditioned, in effect
prohibiting “counter-cyclical deficit spending” (Levitsky and Murillo
2003, 154). Consequently, De la Rúa was forced to implement
unpopular austerity policies that restricted public spending to
balancing the public deficit (Llanos and Margheritis 2006, 94). De la
Rúa compromised with the IMF to lower the public deficit to zero
and to negotiate a “debt swap” (Corrales 2002, 38). Both efforts did
not work out. Owing to the De la Rúa’s broken relations with his
party, his attempts to use tax revenues and spending cuts for lowering
the public deficit failed in Congress (ibid.). Public opinion had started
to harshly criticize De la Rúa’s decisions to adopt austerity policies.
De la Rúa’s stated that he was committed “to pay the debt under all
circumstances” (Schamis 2002, 82), which increased public discontent
since the president seemed to be more worried about reducing the
public debt than to satisfying people’s needs.
26
In 2001 the economy underwent its third consecutive year of
negative growth, the unemployment rate was at a high level again
(18.2%), and the external debt seemed to not have a ceiling ($82,994
million). To make things worse, in March 2001, the FREPASO
leader, Graciela Fernandez Meijie, minister of social-welfare, also
resigned (Schamis 2002, 87). De la Rúa’s weak political capital kept
diminishing. As De la Rúa’s first electoral test approached, street
protests began anew. The provinces were adversely affected by the
economic downturn and executive financial mismanagement. Llanos
(2010, 64) explains that the central government was not even able to
deliver the “minimal monthly sum…of revenue-sharing tax scheme”
to the provinces, unleashing popular outrage. As expected, De la
Rúa’s government did not pass the mid-term elections test. The PJ
won more seats in the lower house at the expense of the Alianza. The
Justicialistas presented a stronger opposition to De la Rúa since there
was no one willing to share the cost “with a government that
everyone else had shunned” (Corrales 2002, 38). Executivelegislative’s prospects for cooperation were extremely unlikely. On
one hand, there were widespread demands for injecting more
government resources into the economy; on the other hand, the
opposition-controlled Congress was unlikely to agree to cutting more
than $4 billion in social spending and salaries in the 2002 budget
(Krauss 2001). Moreover, the PJ even threatened De la Rua by using
the Ley de Acefalia (so-called Acephaly Law), whereby the president of
the Senate may take office when the elected president is declared
“not able” (Llanos 2010, 64.)
In November 2001, as a response to ongoing capital flight,
the Economy Minister limited “bank withdrawals and currency
movements” (Levitsky and Murillo 2003, 154-155). This policy,
known as Corralito (financial corral), proved to be highly unpopular
since it deprived people of their savings (ibid.). In December 2001,
Corralito’s immediate consequences included people protesting on the
streets, riots, looting, piqueteros (road blockers), eighteen people killed,
and mass mobilizations across the country (Llanos 2010, 65). De la
27
Rúa declared a state of siege. On December 19th, the Economy
Minister resigned, and so did the whole cabinet on the next day. De
la Rúa’s last attempt to remain in office was to propose a
“government of national unity” to the Justicialistas, which they
rejected (Schamis 2002, 85). The reasons? Schamis (2002, 85) argues
that when they took De la Rúa’s offer to the UCR leaders, the latter
refused to “take part in any government [De la Rúa] might head.”
After brutal police repressions against protesters in Buenos Aires,
that left several dead, De la Rúa finally resigned on December 20th,
2001, in the middle of his constitutional term. Schamis (2002, 86)
notes:
“Fernando de la Rúa fell in precisely the same
way he had governed: cut off from the larger
political society, at odds with his own party, and
surrounded by a clique of unelected, nonpartisan
advisors, several of whom had no previous
political experience of any kind.”
After analyzing De la Rúa’s presidency, it is possible to
attribute his presidential breakdown to the two structural
determinants of presidebilismo along with street protests. A divided
government, a broken economy, and few but very powerful mass
mobilizations reduced the chances for De la Rúa to stay in office.
Even though he inherited the adverse economic conditions from
Menem’s irresponsible administration, De la Rúa also inherited a
weak electoral and political legacy from the UCR. Nevertheless, the
president himself wiped out the only few political alternatives when
breaking relations, first, with FREPASO, and later with his own
party. The street protests worked as expected, as the last element that
finally led the president to his political demise.
V)
Conclusions
The question what drives presidebilismo? has been partially
answered. Executive-legislative relations and economic conditions are
28
the major determinants of presidential survival. The role of mass
mobilization is influential mostly in the last part of the causal chain.
Street protests tend to appear only once economic conditions have
worsened and perhaps serve to accelerate the presidential ouster.
Whereas executive-legislative relations and economic conditions act
as major structural determinants of presidebilismo, mass mobilization
works as a “precipitant” factor.
The three cases described and analyzed enable us to
understand how the factors that drive presidential stability interact
and then how they might impact on the dependent variable. Under
Alfonsín’s administration, economic crisis and minority government
coincided only at the end of his administration. Only when the
hyperinflation crisis took place did important street protests trigger
Alfonsín’s anticipated departure from office. The case of Menem
provides several examples of how the two structural determinants
can offset each other. His first term was characterized by cooperative
relations between the executive and the legislature. Even though
Menem tended to act unilaterally, which somewhat antagonized the
legislature, parties in Congress did not oppose his initiatives. He had
enough time (2 years) to solve the inflationary crisis without being
threatened even once. Menem’s second term offers more details
about how presidential stability starts to weaken when either the
legislative support or the economy fails. Still, Menem was able to
remain in office, and was almost able to run for a second reelection.
Finally, the case of De la Rúa best represents a “perfect storm” for
presidential survival. Not only were economic woes present as soon
as he took office, he never enjoyed legislative support. In addition,
De la Rua contributed to his own ouster by giving up the only
political resources that he had in Congress: the Alianza.
Further research is needed so as to analyze the role of nonpartisan political actors and unions. Indeed, one common feature
shared by the two non-Peronistas parties was their lack of influence on
sindicatos (Corrales 2002, 34). Political leadership, especially in De la
29
Rúa’s administration, proved to play a major role in the relationship
between the president and his own party/coalition. International
forces also put pressure on and conditioned the set of policies
available for presidents to deal with domestic economic crises. These
three issues have been overlooked in the literature of presidential
stability.
Finally, from a methodological viewpoint, it is important to
understand and take into account the historical sequence of events
that help us to explain presidebilismo. For instance, without considering
the fiscal irresponsibility and economic reforms that took place
during Menem’s administration, it is not possible to comprehend why
and how De la Rúa was so extremely conditioned when he took
office; and also why and how Menem did not fall. Quantitative
studies basically tend to assume that the political system, the
economy, and people’s beliefs reboot every calendar year or with the
start of a new government. This is where the strength of path
dependence approach rests: in explaining how past events still can
affect future outcomes.
Reference
Alvarez, Michael E., and Leiv Marsteintredet. 2010. “Presidential
Democracies Breakdowns in Latin America: Similar Causes,
Different Outcomes.” In Presidential Breakdowns in
Latin America:
Causes and Outcomes of Executive Instability in Developing Democracies,
Eds. Maria Llanos and Leiv Marsteintredet. New York, NY: Palgrave
McMillan.
30
Amorin Neto, Octavio. 2006. “The Presidential Calculus Executive
Policy Making and Cabinet
Formation
in
the
Americas.”
Comparative Political Studies, 39(4): 415-440.
Andreassi, Celina. 2011. “The History of the UCR (part II).”
Argentina Independent, posted September
26th.
www.argentinaindependent.com/currentaffairs/analysis.
Baumgartner, Jody C., and Naoko Kada. (Eds.). 2003. Checking
Executive Power: Presidential
Impeachment in Comparative Perspective.
Westport, CT: Praeger.
Cheibub, Jose A. 2002. “Minority Governments, Deadlock
Situations, and the Survival of Presidential Democracies.” Comparative
Political Studies, 35: 284-312.
Cheresky, Isidoro 1990. “Argentina. Un paso en la consolidación
Democrática: Elecciones
Presidenciales
con
Alternancia
Política.” Revista Mexicana de Sociologia, 52(4): 49-68.
Corrales, Javier. 2002. “The Politics of Argentina's Meltdown.” World
Policy Journal, 19(3): 29- 42.
Deheza, Grace I. 1998. “Gobiernos de Coalición en el Sistema
Presidencial: América del Sur.” In
El Presidencialismo Renovado:
Institucionalismo y Cambio Político en América Latina,
eds.
Dieter
Nohlen and Mario Fernández. Caracas, Venezuela: Nueva Sociedad.
Ferreira Rubio, Delia, and Mateo Goretti. 1998. “When the President
Governs Alone: The Decretazo in
Argentina,
1989-93.”
In
Executive Decree Authority, Eds. Carey, John M. and
Matthew
S.
Shugart. Cambridge, UK: Cambridge University Press.
31
Hochstetler, K. 2006. “Rethinking Presidentialism: Challenges and
Presidential Falls in South
America.” Comparative Politics, 38(4):
401-418.
Hochstetler, Kathryn, and Margaret E Edwards. 2009. “Failed
Presidencies: Identifying and Explaining a South American
anomaly.” Journal of Politics in Latin America, 1(2): 31- 57.
Kim, Young Hun, and Donna Bahry. 2008. “Interrupted Presidencies
in Third Wave Democracies.” The Journal of Politics, 70(3): 807-822.
Krauss, Clifford. 2001. “Reeling from Riots, Argentina Declares a
State of Siege.” New York
Times,
posted
December 20th.
www.nytimes.com/2001/12/20/international/20ARGE.html
Levistky, Steven, and Maria V. Murillo. 2003. “Argentina Weathers
the Storm.” Journal of Democracy, 14(4): 152-166.
Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and
Performance in Thirty- six Countries. New Haven, CT: Yale University
Press.
Linz, Juan J. 1990. “The Perils of Presidentialism.” Journal of
Democracy, 1(1): 51-69.
Linz, Juan J. 1994. “Presidential or Parliamentary Democracy: Does it
Make a Difference?” In
The Failure of Presidential Democracy, eds.
Linz, Juan J. and Arturo Valenzuela. Baltimore, MA: The Johns
Hopkins University Press.
Llanos, Maria. and Ana Margheritis. 2006. “Why do Presidents Fail?
Political Leadership and
the Argentine Crisis (1999-2001).”
Studies in Comparative International Development, 40(4): 77-103.
Llanos, Maria. 2010. “Presidential Breakdowns in Argentina.” In
Presidential Breakdowns in
Latin America: Causes and Outcomes of
32
Executive Instability in Developing Democracies, Eds. Llanos, Maria and
Leiv Marsteintredet. New York, NY: Palgrave McMillan.
Llanos, Maria, and Leiv Marsteintredet, eds. 2010. Presidential
Breakdowns in Latin America: Causes and Outcomes of Executive
Instability in Developing Democracies. New York, NY:
Palgrave
McMillan.
Mahoney, James, and Villegas, Celso M. 2008. Historical Enquiry and
Comparative Politics. In
The Oxford Handbook of Comparative
Politics, Eds. Boix, Carles and Susan C. Stokes. Oxford, NY: Oxford
University Press.
Mainwaring, Scott, and Matthew S. Shugart. 1997. “Presidentialism
and the Party System.” In
Presidentialism and Democracy in Latin
America, Eds. Scott Mainwaring and Matthew S. Shugart. Cambridge,
UK: Cambridge University Press.
Ministerio del Interior de Argentina. 2008. Historia electoral Argentina
(1912-2007). Buenos Aires: Presidencia de la Nacion Argentina.
Montalbano, William D. 1985. “Argentine General Strike, Protests
over IMF Role Challenge
Alfonsín.” Los Angeles Times. Posted
th
May 24 . http://articles.latimes.com/1985-0524/news/mn17074_1_imf-agreement
Mustapic, Ana M. 2000. “’Oficialistas y diputados’: Las Relaciones
Ejecutivo-Legislativo en la
Argentina.” Desarrollo Económico,
39(156): 571-595.
Mustapic, Ana M. 2002. “Oscillating Relations: President and
Congress in Argentina.” In
Legislative Politics in Latin America, Eds.
Scott Morgenstern and Benito Nacif, and
Peter
Lange.
Cambridge, UK: Cambridge University Press.
33
Mustapic, Ana M. 2010. “Presidentialism and Early Exits: The Role
of Congress.” In
Presidential breakdowns in Latin America: Causes
and outcomes of executive instability
in developing democracies, Eds.
Maria Llanos and Leiv Marsteintredet. New York, NY:
Palgrave McMillan.
Negretto, Gabriel N. 2006. “Minority Presidents and Democratic
Performance in Latin America.” Latin American Politics & Society,
48(3): 63-92.
Pérez-Liñán, Anibal. 2007. Presidential Impeachment and the New Political
Instability in Latin
America. Cambridge, UK: Cambridge
University Press.
Przeworski, A. and Limongi, F. 1997. “Modernization: Theories and
Facts.” World Politics,
49(2): 155-183.
Schamis, Hector E. 2002. “Argentina: Crisis and Democratic
Consolidation.” Journal of
Democracy, 13(2): 81-94.
Schuster, F.L., Perez, G.J., Pereyra, S., Armesto, M., Armelino, M.,
Garcia, A.,…Zipcioglu, P.
(2006). Transformaciones de la protesta
social en Argentina, 1989-2003. Grupo de
Estudios
Sobre
Protesta Social y Acción Colectiva. Universidad de Buenos Aires.
Retrieved from <vanina-ue.academia.edu>
Thibaut, Bernhard. 1998. “El Gobierno de la Democracia
Presidencial: Argentina, Brasil, Chile y Uruguay en una Perspectiva
Comparada.” In El Presidencialismo Renovado: Institucionalismo
y
Cambio Político en América Latina, Eds. Dieter Nohlen and Mario
Fernández. Caracas, Venezuela: Nueva Sociedad.
Valenzuela, Arturo. 2004. “Latin American Presidencies Interrupted.”
Journal of Democracy,
15(9): 6-19.
34
World Bank. (2011). World dataBank.World Development Indicators and
Global Development
Finance. Retrieved November 29th, 2010 from
databank.worldbank.org
35
Download